基于经验模态分解法的光学条纹图像处理研究进展

条纹图处理是光学测量技术中一个非常重要的步骤。从早期的经典傅里叶变换,到随后引入局部分析能力的窗口傅里叶变换、小波变换、S变换,再到近些年变分模型分解、经验模态分解(EMD)等,条纹图处理技术经历了长足的研究和发展历程。在这些优秀的技术中,EMD算法由于具有较强的自适应性和复杂信号处理能力而在近些年受到一定的关注。文中结合条纹图处理的关键内容和发展历程,重点分析和总结了EMD算法及其应用于条纹图处理的关键问题和研究进展,指出了该技术尚存的技术难点和主要问题,为相关技术的发展提供了理论和实践的参考。

Researches of optical fringe pattern analysis based on EMD algorithms

Fringe pattern analysis is an important content for optical measurement. The fringe pattern analysis have been researched for a long time, which have introduced the Fourier transform, the windowed Fourier transform, the wavelet transform, the S transform, the recent variable mode decomposition and empirical mode decomposition (EMD) and etc. Compared with other methods, EMD has strong abilities being adaptive and suitable for complex signals, so it gets more attentions in recent years. In this paper, combined with the key contents of fringe pattern analysis, the EMD algorithms and their applications in fringe pattern processing were reviewed. The remaining key issues and technical difficulties were concluded, which were expected to provide some reference for future researches in relevant fields.